Hello again,
So, my second issue with my 2866, sorry for asking, but I am learning at this, and I don't know where else to turn; this is new to me when I have to "go under the hood", as they say.
Plus, not being well doesn't help.

So, I had an issue with my old 2860, so I upgraded to 2866.
I still had an issue after loading the settings file from the 2860 into the 2866.
We eventually traced the slow speed of the 2866 to these settings, "hardware acceleration is enabled, QOS and bandwidth management is disabled, data flow monitor is turned off".
However, before finding this, I started changing the MTU values
Now I don't know what the defaults are for the 2866 or if there are default values.
If there are defaults, would they be for the connection, connection provider or the device (2866)?
We have two incoming connections, BT business FTTC PPPoE & VM Voom Ethernet, to their DOCSIS cable modem they provide, their brand, but I think it is a Hitron Chita.
Now I can't change the MTU in the VM unit. They manage that, and as the connection from the 2866 to the VM unit is Ethernet, my research says that the MTU should be 1500 for ethernet.
My computer's ethernet ports are set to 1500.
The BT is down at 1492 because I found that as the default for a PPPoE connection.
However, if I run a test as advised on many websites, I get much lower values for the MTU than these.
Plus, they vary up and down depending on when I test them and where to.
Should I change/set the MTUs?
If so, has anyone got any good recommendations, please?

Many thanks in advance.

Hi

My WAN2 is set to 1492. I have a 5G router plugged in to it... It seems to work very well.

On another note, VDSL seems to be a complete disaster on the 2866! I also upgraded from 2860 and have lost half my bandwidth due to the VDSL line resetting every few hours! What firmware are you on? I have web filtering, so Draytek support put me on 4.4.2 RC10 as 4.4.1 has a known issue with reaching the Cyren servers! Nightmare!

If you VM Modem is in Modem mode (as opposed to router mode) then the MTU will be set from the Vigor rather than the modem, I believe. I had this set up many years ago, but now have backup using 5G router...

Thanks

Simon

@Simon - out of interest what is the build date of that RC10 firmware you have please ?

I found that my 2866Lac also seemed to cause lower line speed after moving from a 2862Lac. However, once all settings were in-place I found that it became pretty stable - and speeded back up - as long as I left it alone rather than messing with the config all the time.
Current uptime on mine is now 1833 hours - VDSL has been up for 391 hours and the LTE for 986.

But I can almost guarantee it'll crash and reboot if I go look at the DSL Status page under Diagnostics!

Thank you for the post, Simon,

I have had to split this into two to give all the info.
Part 1.

We have BT business on our WAN1; PPPoE and FTTC.
It's the fastest connection we can have where we live via BT or OpenReach architecture to the premises without having a new physical line fitted as a leased line or similar.
It is on a static IP.
I am using 4.4.1_BT firmware in the 2866.
Now we get a reasonable and steady connection through the BT, but it is underutilised as I was running my incoming web access and VPN across it, but we have had internal problems with other unrelated parts of our systems, which means that we are without that aspect now, that is the next project after sorting out the MTU and a few other things, I let my network management tasks get behind, then we had a corruption issue, didn't lose any data thank goodness but, loads of sorting out to do and I am working through it systematically so I don't have to duplicate my work, and it's taking me a while.
Having terminal cancer slows you down somewhat! 🤣

If I do a "test" through the "Details" page PPPoE tab and then say do a "detect" to the Google DNS servers the best MTU I can get is around 1340, which seems extremely low for 1GB wired Ethernet to the 2866, OK it goes through a couple of D-ink switches down Cat5E/Cat6 copper and a fibre link between the switches, but that should affect speed rather than packet size?
I don't recall anywhere I have changed MTU in the D-Link units, and we have had issues in various places in the network, including on one wired PC connected straight to port 6 on the 2866 unit.

Having tried 8.8.8.8 this morning I get 1492 back, which is apparently, the "default" MTU for a PPPoE connection, so this "should" now be, OK? 🤷‍♂️

Going on to the VM connection, this is a business line also, but it is their wired system DOCSIS or whatever, it is a dedicated coax coming in.
We have I believe a Hitron Chita modem unit, which is in modem mode as we have static IPs.
A lot of the VM unit settings are not accessible as it is not in modem/switch/hub/wi-fi/router mode, it is just a modem to convert our ethernet from the 2866 into "VM speak" for sending down their network.
Apparently, this modem is only used for the high speed 1GB connections. 🤷‍♂️

End of Part 1.

Part 2.

Speed on the VM seems fine.

I have just done an Ookla speedtest though the app on my Win 10 pro x64 i9 Lenovo laptop wired to a D-link DGS-1210-24 which has a parallel fibre connection to a Netgear M4100-12GF which is then ethernet to the 2866 and ethernet to the VM modem.

All ethernet are indicating at 1 gig connections.

To Swish Fibre in London 13ms ping, 898.24Mbs down, 102.53 Mbps up.

Which I don't think is too bad as we were running two media streaming devices, playing radio at the same time when doing the speed test, plus the running apps across the network, I am connected to another PC on my local network remotely from the laptop, and there will be other unseen traffic, I'm sure.

This all came up because my son who does a lot of PC gaming couldn't log into one of his games, and the error for this is common and has been around for years, but the providers have not generated a fix, and it's another thing like MTU where there are lots of ideas on the internet, and some help some people and others help other people, there is no definitive answer.
One of the things that came up in our hunt for his of in problem was MTU sizes, so that is why we went down this path.
We have somehow fixed his access, but I'm not sure if I should persevere with the MTU and if so what to do/set them at!

We had terrible issues when they did the upgrade, no connection at all through the unit, we were told that the 1GB simply didn't work with static IPs, but, I dug my heels in and they did sort it at "level 3" support or above, but, it was a massive change as in they had to re-create our VM accounts completely, new account number log in everything, apart from IPs which I did ask if we could keep, though it would have been OK to have new ones.

As yet, I have stayed away from a 4/5G connection, between the two we have, it's not been needed, and IF push comes to shove, I might be able to do something with my mobile, but it's never come to that.

This may sound extravagant, but we ran the business from home, and it was data heavy, I travelled a lot so fast access to cloud data and being able to VPN in from all over the world was important to me.
Even with the cancer, I am trying to keep the business going in case they find a cure or a treatment that helps me enough to get working again, so I am using the time to re-configure everything and simplify stuff so if I do go my family can easily re-configure things to simplify them and will still be able to keep all our personal data.

This is really at the edge of my network knowledge to be honest and if anyone can help point me in the way of learning resources rather than giving me the answers outright, I will do the legwork to learn. (Though to be honest I would appreciate the definitive answer if anyone has it, because time is precious to me with my disease).

Thanks in advance for any help that I can have, and thanks again to SimonH for kicking off the answers for me.

End

I can't get to the build date as I am now on some beta test firmware instead to see if it fares any better! It's all been a real pain... I also had the other day where with RC10 my router crashed and rebooted when I tried to go to the Online Status page!

Mine seems to lose sync every 5 to 10 hours. Hence why i have lost so much of my bandwidth!

ianfretwell wrote:
@Simon - out of interest what is the build date of that RC10 firmware you have please ?

I found that my 2866Lac also seemed to cause lower line speed after moving from a 2862Lac. However, once all settings were in-place I found that it became pretty stable - and speeded back up - as long as I left it alone rather than messing with the config all the time.
Current uptime on mine is now 1833 hours - VDSL has been up for 391 hours and the LTE for 986.

But I can almost guarantee it'll crash and reboot if I go look at the DSL Status page under Diagnostics!